Project number: MT-11-06

Funding source: Caltrans

Total cost: $89,910

Performance period: August 2010 to July 2011

The purpose of this research is to examine the relationship between accessibility and the growth of employment centers in order to improve our understanding of how transportation investments influence the spatial organization of metropolitan areas. Although research on the existence of employment centers - concentrations of employment outside the tradition downtown - is extensive, we have little understanding of how these centers emerge and grow, and what role transportation access may play in this process. Studies of employment centers are limited by data availability: there is no publicly available source for reliable, highly detailed and disaggregate employment data. For this research we used the National Establishment
Time-Series (NETS) Database, a proprietary database that consists of time series establishment level data for the entire US for the years 1990 through 2009. The NETS data are highly detailed; they provide information on business type, size, location, corporate structure, etc. Despite the richness of the database, however, we discovered that it is not comparable to any other data source, raising serious questions of validity. We devote one chapter of this report to an assessment of the quality and reliability of the NETS data.

We used the NETS data to identify employment centers in California's four largest metro areas (Los Angeles, San Diego, San Francisco and Sacramento) and examine their evolution over the time period in terms of number of firms, employment, and other factors. We find that the four metro areas have quite different urban structures. Los Angeles and San Francisco are highly polycentric, while San Diego and Sacramento are less so, and these characteristics are persistent over the time period of the study. We conduct formal tests of polycentricity, and find that all confirm polycentric urban form. However, the influence of centers outside the CBD is weakest for Sacramento. Consistent with prior studies, the extent of polycentricity is related to metropolitan size. We estimated models of employment center growth as a function of accessibility. We estimate one model for the Los Angeles region, and another model with a pooled sample of Los Angeles and San Francisco. We find that center density is the most consistent factor associated with employment center growth: growth is negatively related to employment density. Results on our access measures - access to airports and freeways, access to labor - are mixed. We attribute our results to the unreliability of the NETS data when used at a highly disaggregate geographic scale, the relatively slow growth that took place over the study period, and the overall maturity of urban structure in Los Angeles and San Francisco. All employment centers have relatively high access to the regional transport system and labor force, hence differences in growth rates are due to more place specific factors.

Our results suggest the need for both more reliable data sources for detailed, spatially disaggregate data and comparable studies of metro areas outside California. The role of transport and labor force accessibility in employment center growth remains uncertain.